Bulk SmCo5/a-Fe nanocomposite permanent magnets fabricatedby mould-free Joule-heating compaction
Chuanbing Rong,1,a) Ying Zhang,1,2 Narayan Poudyal,1 Dapeng Wang,1 M. J. Kramer,2
and J. Ping Liu1,b)
1Department of Physics, University of Texas at Arlington, Arlington, Texas 76019, USA2Division of Materials Science and Engineering, Ames Laboratory, USDOE, Iowa State University,Ames, Iowa 50011, USA
(Presented 18 November 2010; received 22 September 2010; accepted 21 December 2010;
published online 7 April 2011)
Bulk SmCo5/a-Fe nanocomposite magnets have been prepared using a Joule-heating compaction
technique. Nearly fully dense bulk magnets are obtained by compacting the milled powders under
a pressure of 2 GPa at temperatures above 400 �C. Structural analysis shows that the grain size of
both the SmCo5 and the a-Fe phases is in the range of 10 to 15 nm when the compaction
temperature is lower than 500 �C, which ensures effective interphase exchange coupling. A further
increase in compaction temperature leads to significant grain growth and deterioration of magnetic
properties. A maximum energy product of about 18.5 MGOe was obtained in the bulk SmCo5/a-Fe
nanocomposite magnets, which is 90% higher than that of the single-phase counterpart prepared
under the same conditions. VC 2011 American Institute of Physics. [doi:10.1063/1.3563098]
I. INTRODUCTION
Nanocomposites consisting of magnetically hard and
magnetically soft phases are being extensively investigated
due to their potentially superhigh maximum energy products
[(BH)max]. The high energy products originate from the
interphase exchange coupling between the neighboring hard
and soft magnetic phases.1–3 The prerequisite for effective
exchange coupling is a small grain size of the soft magnetic
phase, with the critical dimension estimated to be on the
order of 10 nm.4 The experimental nanocomposite materials
are typically in the form of low dimensional powder particles
or thin films.5–9 For application purposes, the fabrication of
bulk nanocomposite magnets is essential. However, there are
major obstacles to producing bulk nanocomposite magnets.
Conventional sintering and hot-pressing methods, which are
used to produce single-phase microcrystalline permanent
magnets, are ineffective in producing bulk nanocomposite
magnets, because those techniques require extensive heat
treatments, which lead to excessive grain growth. Recent
attempts to fabricate bulk nanostructured magnets by using
unconventional and expensive compaction techniques,
including spark plasma sintering (SPS),10,11 dynamic com-
paction,12 and warm compaction,13 have achieved grain size
control in the nanocomposite systems. In this paper, we
report our recent work in an alternative methodology for fab-
ricating bulk nanocomposite magnets by using a Joule-heat-
ing compaction technique.
II. EXPERIMENT
The raw materials of the nanocomposites were prepared
by high energy ball milling the hard magnetic phase SmCo5
particles and the soft magnetic phase a-Fe particles. The
weight ratio of powder to ball is around 1:30, and the milling-
time is 4 h. The as-milled powders contain an amorphous
SmCo matrix and a bcc structured soft magnetic phase. The
as-milled powders were sealed in a container with �60% ini-
tial density followed by the Joule-heating compaction under a
pressure of 2 GPa using a modified QUICKpress apparatus.
For comparison, the single-phase SmCo was also prepared by
the same process. The schematic view of the Joule-heating
compaction set-up is shown in Fig. 1. By changing the
applied current through the samples, the heating of the sam-
ples is adjustable from room temperature to 650 �C. The den-
sity of the compacts was measured by the Archimedes
method and by measuring the mass and the volume of regular
shaped bulks. The Vickers hardness was measured using a
LECO Microindentation Hardness Tester LM247AT. The
morphology and crystalline structure were characterized by
scanning electron microscopy, transmission electron micros-
copy (TEM), energy-filtered TEM (EFTEM), and x-ray dif-
fraction (XRD) using Cu Ka radiation. Magnetic properties
were measured in a superconducting quantum interference de-
vice magnetometer with a maximum applied field of 70 kOe.
III. RESULTS AND DISCUSSION
The Joule-heating compaction (JHC) in this work was per-
formed on an in-house built apparatus. The fixed ac current is
directly passed through the loading rams as well as the powder
compacts, as shown in Fig. 1(a). The generated Joule heat, i.e.,
resistance heating from the sample, internally heats the sample.
This method is different from conventional hot pressing, in
which the heat is provided by external heating elements, such
as graphite furnaces [as shown in Fig. 1(b)]. Our facility is simi-
lar to spark plasma sintering, also known as pulsed electric cur-
rent sintering.10,11 However, there are several differences
between JHC and SPS: (1) SPS adopts a pulsed dc current,
a)Electronic mail: [email protected])Electronic mail: [email protected].
0021-8979/2011/109(7)/07A735/3/$30.00 VC 2011 American Institute of Physics109, 07A735-1
JOURNAL OF APPLIED PHYSICS 109, 07A735 (2011)
Author complimentary copy. Redistribution subject to AIP license or copyright, see http://jap.aip.org/jap/copyright.jsp
whereas the JHC in our work uses a stable ac current; and (2)
SPS normally uses graphite punches to push the sample,
whereas JHC uses a tool steel piston. The tool steel provides
higher pressure (up to 2 GPa in this work) than graphite. In
addition, the steel has lower resistance than graphite, and thus
most of the heat is generated from the sample itself in the JHC
technique. This facility has the ability to heat samples at a rate
of around 300 �C/min. The fast heating benefits the grain size
and nanostructure controlling in the compaction processes for
the preparation of magnetic hard/soft nanocomposite magnets,
in which the grain size of the magnetically soft phase is crit-
ically important. We will demonstrate the experimental results
based on the SmCo5/a-Fe nanocomposite system in the follow-
ing paragraphs.
Figure 2 shows the dependence of the density of the
bulk samples on the compaction temperature, Tcp. One can
see that the samples compacted at room temperature have a
density of only about 7.4 g/cm3, which is about 88% of the
theoretical value (8.4 g/cm3). The density increases monot-
onically with increasing Tcp and reaches saturation when Tcp
is above 450 to 500 �C. This is attributed to the fact that both
SmCo5 and Fe become more compressible upon heating to
high temperatures. The interphase diffusion at high tempera-
tures may also help the consolidation, as reported in our pre-
vious work.13 Figure 2 also gives the dependence of the
Vickers hardness on Tcp with a load of 500 gs. The hardness
increases fast with Tcp up to 500 �C, and then the increase
slows down. This also indicates that a nearly full density is
obtained when Tcp is above 500 �C.
Figure 3(a) shows the XRD patterns of the bulk
SmCo5þ 20 wt.% a-Fe magnets compacted from amorphous
powders at different temperatures. Only broadened peaks for
the bcc a-Fe phase were observed when Tcp was lower than
400 �C. It is obvious that the SmCo matrix is amorphous. The
crystalline SmCo5 phase starts to form when Tcp is higher than
500 �C because the crystallization temperature of amorphous
SmCo is about 470 �C, according to differential scanning calo-
rimetry measurements. Compaction at 540 �C leads to the for-
mation of a well-developed SmCo5 phase; however, further
increases of Tcp result in sharper peaks of both the SmCo5 and
the a-Fe phases, indicating significant grain growth. A quanti-
tative analysis was performed using the Williamson–Hall
method13,14 [Fig. 3(b)]. The measured a-Fe grain size is about
6 nm for the bulk compacted at room temperature. With
increasing Tcp, the grain size of the a-Fe phase remained con-
stant to �400 �C and started to increase with Tcp above
450 �C. Fortunately, the grain growth was not excessive, and
the grain size is less than 15 nm if Tcp is kept under 550 �C,
while significant grain growth is observed when Tcp is higher
than 550 �C. It should be noted that the Sm–Co phase was
amorphous when Tcp was lower than 470 �C, so there were no
data for grain size below that temperature. For comparison,
the grain size of single-phase SmCo magnets prepared under
same conditions is also given in Fig. 3(b).
TEM micrographs allow for direct observation of the
morphology of the nanostructured composites. Figures 4(a)
and 4(b) show the morphology and Fe element distribution
of the bulk sample compacted from amorphous powders at
room temperature. A higher brightness in the EFTEM image
corresponds to a higher Fe concentration in the Fe distribu-
tion map. One can see that the Fe grain size is as small as 5
nm and the grain distribution is homogenous. It should be
noted that codiffusion took place between the Sm–Co and Fe
phases, so, actually, the magnetically soft phase contains Co,
similar to the bulk nanocomposite magnet prepared by warm
compaction.15,16 As Tcp increases to 540 �C, the grain size of
FIG. 2. (Color online) The dependence of density and Vicker hardness on
Tcp of the bulk magnets.
FIG. 3. (Color online) (a) XRD patterns of the bulk magnets compacted at
different temperatures. (b) The dependence of the grain size of both the
SmCo5 and the a-Fe phases on Tcp of the bulk magnets.
FIG. 1. (Color online) Schematics of (a) Joule-heating compaction and (b)
conventional heating facilities.
07A735-2 Rong et al. J. Appl. Phys. 109, 07A735 (2011)
Author complimentary copy. Redistribution subject to AIP license or copyright, see http://jap.aip.org/jap/copyright.jsp
the a-Fe phase increases to 8 to 15 nm, as shown in Figs.
4(c) and 4(d); this is still within the effective exchange-
coupling length. The magnetically soft phase is also well
distributed in the Sm–Co matrix. However, when the com-
paction temperature was increased to 640 �C, the overall
grain size significantly increased to 20 to 25 nm, as shown in
the TEM image of Fig. 4(e). The TEM analysis confirms the
XRD results quite well. For comparison, Fig. 4(f) shows the
TEM image of the SmCo5/Fe magnets compacted by con-
ventional hot compaction at 640 �C for 10 min under a pres-
sure of 2 GPa. The grain size (30 to 40 nm) of the
conventionally compacted magnets is significantly larger
than that of the JHC compacted magnets (20 to 25 nm). The
smaller grain size of the JHC compacted sample may be due
to the fast and internal heating and cooling.
Figure 5(a) shows the demagnetization curves of the
bulk magnets compacted at different temperatures. Similar
demagnetization curves with low coercivity were observed
for the samples compacted at 20 and 400 �C, because Tcp is
lower than the crystallization temperature. The crystalliza-
tion process starts in the 500 �C compacted samples, while a
high enough coercivity is obtained only when Tcp is higher
than 540 �C. The squareness of the demagnetization curves
of the 540 �C compacted samples is attributed to the fine
grain size below 15 nm, which is close to the effective
exchange-coupling length, i.e., twice the domain wall width.
A kink was observed in the 640 �C compacted magnets, most
likely due to the grain growth and insufficient interphase
exchange coupling. Figure 5(b) shows the dependence of the
coercivity (Hc) and the energy product (BH)max on Tcp of the
bulk nanocomposite magnets. The (BH)max of single-phase
magnets is also given for comparison. One can see that Hc is
only about 1 kOe when Tcp is lower than 450 �C, and thus
(BH)max is low. The coercivity reaches a maximum when Tcp
is about 540 �C. The maximum (BH)max of about 18.5
MGOe—which is 90% higher than that of SmCo5 single-
phase nanocrystalline magnets—was also obtained in the
540 �C compacted samples.7 A further increase of Tcp leads
to a decrease in (BH)max due to the grain growth.
IV. CONCLUSIONS
In summary, a Joule-heating compaction technique has
been developed to produce fully dense bulk SmCo5/a-Fe
nanocomposite magnets under high pressure at moderate
temperatures. The grain size of the magnetic soft phase can
be controlled under 15 nm, which ensures effective exchange
coupling between the hard and the soft phases. A maximum
energy product of 18.5 MGOe has been obtained, which is
90% higher than that of the single-phase counterpart. It was
also observed that the grain size of the Joule-heating com-
pacted bulks is smaller than that of the conventional hot
pressed bulks prepared under similar conditions.
ACKNOWLEDGMENTS
This work has been supported in part by the U.S. Office of
Naval Research/MURI project under grant N00014-05-1-0497
and DARPA/ARO under grant W911NF-08-1-0249. The mi-
croscopy was performed at the Ames laboratory, which is sup-
ported in part by the U.S. Department of Energy, Office of
Basic Energy Science under contract DE-AC02-07CH11358.
1R. Skomski and J. M. D. Coey, Phys. Rev. B 48, 15812 (1993).2T. Schrefl, H. Kronmuller, and J. Fidler, J. Magn. Magn. Mater. 127, L273
(1993).3C. B. Rong, H. W. Zhang, R. J. Chen, S. L. He, and B. G. Shen, J. Magn.
Magn. Mater. 302, 126 (2006).4E. F. Kneller and R. Hawig, IEEE Trans. Magn. 27, 3588 (1991).5J. Bauer, M. Seeger, A. Zern, and H. Kronmuller, J. Appl. Phys. 80, 1667
(1996).6G. C. Hadjipanayis, J. Magn. Magn. Mater. 200, 373 (1999).7J. Zhang, S. Y. Zhang, H. W. Zhang, and B. G. Shen, J. Appl. Phys. 89,
5601 (2001).8Z. C. Wang, H. A. Davis, S. Z. Zhou, M. C. Zheng, and Y. Qiao, J. Appl.
Phys. 91, 7884 (2002).9H. Zeng, J. Li, J. P. Liu, Z. L. Wang, and S. Sun, Nature 420, 395 (2002).
10M. Yue, J. Zhang, M. Tian, and X. B. Liu, J. Appl. Phys. 99, 08B502
(2006).11C. B. Rong, V. Nandwana, N. Poudyal, J. P. Liu, T. Saito, Y. Q. Wu, and
M. J. Kramer, J. Appl. Phys. 101, 09K515 (2007).12C. D. Dai, D. Eakins, N. Thadhani, and J. P. Liu, Appl. Phys. Lett. 90,
071911 (2007).13C. B. Rong, V. Nandwana, N. Poudyal, J. P. Liu, M. E. Kozlov, R. H.
Baughman, Y. Ding, and Z. L. Wang, J. Appl. Phys. 101, 023908 (2007).14K. Williamson and W. H. Hall, Acta Metall. 1, 22 (1953).15C. B. Rong, Y. Zhang, N. Poudyal, X. Y. Xiong, M. J. Kramer, and J. P.
Liu, Appl. Phys. Lett. 96, 102513 (2010).16Y. Zhang, M. J. Kramer, C. B. Rong, and J. P. Liu, Appl. Phys. Lett. 97,
032506 (2010).
FIG. 4. (a) TEM image and (b) Fe element map of a room-temperature com-
pacted magnet. (c) TEM image and (d) Fe element map of a 540�C com-
pacted magnet. (e) TEM image of a 640 �C compacted magnet. A TEM
image of the magnet compacted at 640 �C by traditional methods is shown
in (f) for comparison.
FIG. 5. (Color online) (a) Demagnetization curves of the bulk samples com-
pacted at different temperatures. (b) The dependence of Hc and (BH)max on
compaction temperatures.
07A735-3 Rong et al. J. Appl. Phys. 109, 07A735 (2011)
Author complimentary copy. Redistribution subject to AIP license or copyright, see http://jap.aip.org/jap/copyright.jsp